Sur cette page
Time prediction
Based on a recent performance, project your target times over other distances. Accuracy ±2-5% at close distances.
The Riegel formula
The reference formula for predicting running times was proposed by Peter Riegel in 1977:
T2 = T1 × (D2 / D1)^1.06
where T1 is your known time over distance D1, and T2 your predicted time over distance D2. The exponent 1.06 reflects the fact that we lose a little pace as the distance increases.
Precision depending on distance deviation
- 5K → 10K: ±2% (very accurate).
- 10K → semi: ±3% (precise).
- 10K → marathon: ±5 to 10% (less reliable, especially if you do not have suitable marathon training).
- 5K → marathon: ±10 to 15% (unreliable, projection too far).
Golden rule: The prediction is only valid if you have the specific endurance at the target distance. An excellent 5K time does not guarantee a good marathon without long preparation (long outings of 30-35 km in the training plan).
Why the exponent 1.06?
If the exponent was 1, the average speed would be constant regardless of the distance. In reality, it drops slightly when the distance increases (fatigue, glycogen, thermal management). The exponent 1.06 is a statistical average over large populations of runners.
Variants used by other models:
- Cameron (1998): slightly more precise alternative for elite marathon runners.
- Daniels VDOT: proprietary table used by Daniels Running Formula coaches.
- Race Time Predictor (Garmin/Strava): proprietary models based on GPS + cardio data.
How to improve accuracy for the marathon
The marathon is the distance least well predicted by Riegel for two reasons: the 30ᵉ km wall (glycogen depletion) and specific endurance. To make it more reliable:
- Do at least 3-4 long rides of 30+ km in the 8 weeks before the race.
- Test your marathon pace on a long outing (20 km at planned marathon pace) 3-4 weeks before. If you hold on without suffering: the prediction is credible.
- Add 3-5% to the default Riegel marathon prediction (safety).
Calculation limits
The formula ignores:
- The route profile: elevation, type of terrain.
- Weather conditions: heat (+1% per 5°C above 15°C).
- Your physiological profile: a “fast” runner (strong PCR) will be better in 5K than in marathon; vice versa for an “endurance” runner.
- Your freshness of the day: 5% easy difference depending on sleep and recovery.
Questions fréquentes
Yes. Riegel often overestimates the marathon, especially if your benchmark is a 5K or 10K. Allow +3 to +5% margin if you do not have previous marathon experience.
Ideally a recent semi (< 6 weeks), run hard. A 10K also works but with less precision. A 5K alone is insufficient to project a reliable marathon.
As long as your form remains stable. During periods of progression, retest every 2-3 months. After an outage (vacation, injury), retest before programming.
Because the Riegel formula assumes a linear increase in difficulty with distance. On huge gaps (5K vs marathon), a marathoner can be better than expected in 5K if he also did speed in training.